The Heritage Foundation

Backgrounder #2480 on Taxes

October 21, 2010

October 21, 2010 | Backgrounder on Taxes

Eliminating Tax Expenditures: Beware the Third Wave of Tax Hikes

Key Points

  1. The possibility of repealing various tax provisions called “tax expenditures” is receiving increasing attention in budget and tax policy circles.
  2. An obvious problem with eliminating tax expenditures to raise revenue is that projected outsized budget deficits are due entirely to surging federal spending, not to a sudden dearth of tax revenues. Federal spending has traditionally absorbed about 20 percent of the U.S. economy. Today, it represents nearly 25 percent.
  3. Another problem is that no reliable analysis of tax expenditures is available to guide policymakers. Reports by the Treasury Department and the Joint Committee on Taxation are simply too fundamentally flawed. In these reports, tax expenditures are largely in the eye of the beholder, an analytically deficient approach that is useless as a guide either to tax reform or even tax raising.
  4. When Americans hear “reforming tax expenditures” in the current environment, they should think “tax hike ruse.”

Abstract: The “Obama tax hike”—the expiration at the end of 2010 of tax relief enacted nearly a decade ago—is a major issue facing American taxpayers. Often overlooked is yet another tax hike fermenting in Washington—the elimination of tax expenditures to shrink budget deficits. This hike, too, is misguided: Projected and unsustainable budget deficits are due to surging federal spending, not a lack of tax revenues. However, even for revenue-neutral tax reform, existing tax expenditure analysis is too flawed to provide a reliable guide for policymakers. Heritage Foundation tax policy analyst J. D. Foster explains the uses and troubles with tax expenditures—and why talk of “reforming” them is, in reality, a ruse for tax hikes.

The great Washington tax hike machine is gearing up on many fronts. The immediate grist for the tax hike mill is the “Obama tax hike”—the expiration of the 2001 and 2003 tax relief.[1] A much bolder, more dangerous possibility is that the President would take advantage of a fiscal crisis brought on by spending-driven trillion-dollar budget deficits to foist a value-added tax[2] in addition to existing taxes. A third possibility waiting in the wings and receiving increasing attention is the possibility of repealing various tax provisions called “tax expenditures.”[3] So-defined deductions, exemptions, exclusions, and credits have long garnered ire as bad tax policy in their own right. Where better then to raise tax revenue than by eliminating bad tax policies?

The obvious first problem with eliminating tax expenditures to raise revenue is that outsized budget deficits in years to come are due entirely to surging federal spending, not to a sudden dearth of tax revenues. Federal spending has traditionally absorbed about 20 percent of the U.S. economy. Today, it represents nearly a quarter of the economy, and that figure is projected to decline only slightly through the balance of the decade before it rises steadily back to and past 25 percent as the entitlement spending wave hits with full effect. Washington must act soon to bring the budget deficit under control. Congress and the President need to go on a big-time-spenders diet in the near term while bringing entitlements to heel over the long run; think Jenny Craig for big spenders.

The second problem is that no reliable analysis or listing of tax expenditures is available to guide policymakers even if one wanted to improve the tax code without raising revenues. Reports produced by the Treasury Department and the Joint Committee on Taxation (JCT) are simply too fundamentally flawed to serve as guides. In these reports, tax expenditures are largely in the eye of the beholder, an analytically deficient approach that is useless as a guide either to tax reform or even tax raising.

The fact that serious and inexcusable flaws contained in the federal tax system remain is an issue on which there is sustained, strong agreement across ideological viewpoints: The tax system is impenetrable for the Internal Revenue Service and outside experts alike; it is excessively costly for both the taxpayer and the tax service; and the distortions it imposes on economic decision making cut into economic growth. Tax expenditure analysis should be a useful tool for understanding the operation of the tax system, its interactions with the economy, and for highlighting areas in which substantive, pro-growth, revenue-neutral reforms might be most advantageous. Before tax expenditure analysis can be a useful guide for tax reform, the Treasury Department and the JCT must first reform the tax expenditure analysis.

Defining a Tax Expenditure

Sound tax expenditure analysis begins with a clear, rigorous definition of a pure tax system, a definition stripped of personal biases and political leanings. It should acknowledge that the federal government today imposes a hybrid tax that mixes income and consumed-income tax aspects that create especially difficult questions when attempting to define a tax expenditure.[4] These questions can probably only be answered by presenting two versions of the tax expenditure analysis, one assuming a strict, comprehensive income tax and one assuming a strict, comprehensive consumed-income tax. Each of the baseline tax systems would assume a single rate of tax applied to the first dollar earned.

A tax expenditure is implicitly a tax loophole broadly defined, but it is not always easy to identify a tax loophole. Sometimes it is easy, such as the exclusion for employer-sponsored health insurance. While President Obama personally defended this particular loophole, at more than $1 trillion over 10 years there is little dispute in tax or health care policy circles that this is the largest and perhaps most distorting tax loophole of all. Premiums paid for employer-sponsored health insurance may be made in whole or in part by the employer, but the employee’s other wage and non-wage compensation is reduced accordingly. These premiums reflect employee compensation that would normally be taxed under either an income tax or a consumed-income tax.

Many tax provisions are much more difficult to pigeonhole as a tax expenditure because a loophole must be defined as a violation of a broad principle or theoretical framework. As the JCT observes in its most recent report on the subject, such a consensus theoretical framework is sorely lacking.[5]

The precondition for identifying a loophole is a clear and carefully formulated standard to know what amount of income should be taxed and what amount of tax should be paid. Lacking such a standard, the present identification of a tax expenditure is often a reflection of tradition and manifestation of personal opinion. Traditions and opinions have their place, but unless the tax expenditure exercise rests on a firm analytic foundation, it has little value and may be seriously misleading. The tax expenditure exercise as currently presented by Treasury and the JCT lacks such a firm analytic foundation, a point the JCT acknowledges as it sets out on a somewhat different approach. Ultimately, however, the JCT analysis merely replaces one standard predicated on opinion with another, and so still lacks a defensible analytical foundation (as discussed below).

Curiously, the federal income tax itself fails to provide a theoretical basis for calculating taxable income. Income is defined in law entirely by construction. Nor does the law requiring Treasury to present a tax expenditure analysis provide useful guidance, as it refers to a “normal” income tax without attempting to define the term. All of which leads to the following sensible summation by the JCT of the state of analysis:

As currently applied, tax expenditure analysis is less helpful to policymakers in fashioning tax policy than might otherwise be the case, because the proponents of tax expenditure analysis generally have failed to respond convincingly to the important criticisms leveled against it. Tax expenditure analysis has always been controversial, and there is today a voluminous literature criticizing its premises and implementation as a tool of tax policy.[6]

The clear implication from what the JCT correctly argues thereafter is that if the analysis cannot be performed in a way that is credible and free of extraneous agendas, it has little value to tax policy.

Without question—tax expenditures abound in the current tax code. The exclusion for employer-provided health insurance is a good example, as is the earned-income tax credit, Obama’s Make Work Pay tax credit, and the ethanol tax credit.

In contrast, a tax provision commonly targeted by many professing concern over tax expenditures—the deduction for home mortgage interest— is demonstrably not a tax expenditure in an income tax. As explained below, there is another aspect of the income tax as it relates to housing that gives rise to an enormous and largely unidentified tax expenditure, but the true tax expenditure is unrelated to the home mortgage interest deduction.

A basic principle of tax policy is that when interest income is taxable to the recipient, it should be deductible for the payer. This is true whether the debt is incurred to pay for college, vacation, car, boat, or home. This symmetry of treatment—taxing the interest income, deducting the interest expense—is necessary to avoid creating a tax distortion. Mortgage interest income is taxed to the lender, so the expense should be deductible for the borrower. Therefore, the deduction is not a tax expenditure.

The true housing tax expenditure arises in that the net imputed income from owning a home is not subject to income tax. (Net imputed income is the income one would earn if renting the house to someone else.[7]) The exclusion of net imputed income is not a policy intended to encourage home ownership, though that surely is a consequence. It is, rather, a simple acknowledgement that it is nearly impossible to calculate net imputed income in any given year, since there is no transaction to record.

Traditional Analysis

Traditional tax expenditure analysis performed by the Treasury Department and the JCT for many years adopted a stylized version of a “normal” income tax and defined as a tax expenditure any provision that lowered tax liability below what would arise under that normal income tax. Many of the oddities of these analyses reflected these stylized choices, which largely reflected the views of Stanley Surrey, the original proponent of tax expenditure analysis. Some examples of these anomalies are:

  • With its fixation on tax loopholes operating in lieu of explicit expenditures, the traditional analysis ignored entirely those provisions and circumstances in which the tax system over-taxed income;
  • It was also silent on those areas in which there was obvious double taxation, such as the corporate income tax;
  • It incorrectly assumed an income tax on the foreign-source income of U.S. citizens;
  • It assumed the income tax should exclude some amount of income for low-income filers; and
  • In an entirely arbitrary adoption of Surrey’s views, it assumed an income tax would have a progressive tax rate schedule.

Not all of the flaws in the traditional tax expenditure analysis can be attributed to the confusion regarding a normal income tax, however. In some cases, it was a simple failing of analysis and a refusal to part with past practices. As discussed, the insistence on defining the home mortgage deduction as a tax expenditure is a good example.

To be sure, traditional tax expenditure analysis is often correct. The tax credits appearing with increasing frequency in the tax code are shown to be tax expenditures, as is the biggest tax expenditure of all—the exclusion for employer-sponsored health insurance. Problems arise for policymakers attempting to use the tax expenditure analysis as a guide, however, because properly defined tax expenditures are indistinguishably intermingled with provisions that are proper to the income tax.

The New JCT Analysis

In its most recent analysis, the Joint Committee on Taxation acknowledged the fatal flaws of the traditional approach and devised another.[8] The JCT should be commended for thinking creatively about this long-standing analytical problem and for proposing an alternative that breaks sharply with tradition. Unfortunately, the JCT approach is ultimately as flawed as the traditional approach.

The new JCT approach is to examine the tax code in search of general rules applied to the code itself, and then to define a tax expenditure as a departure from a general rule. This merely replaces the general rules supposedly reflected in whatever passes for a “normal” income tax with the general rules manifested in the tax code itself. The JCT approach raises a number of issues beginning with the simple question: Who is to say what is a general rule, other than the JCT itself?

Another issue arising from the JCT approach: What if the JCT’s rules vary from those prescribed by a comprehensive income tax? So much progress has been made in eliminating the tax bias against saving through such devices as IRAs and 401(k)s that the income tax appears more like a consumed-income tax. The general rule then appears to be that retirement saving is tax deferred, while instances in which tax deferral is unavailable reflect tax surcharges. On this important and fundamental issue the general rule is very much open to debate.

Yet another issue arises when the tax code changes sufficiently over time that the interpretation of the general rule must change. Is the definition of tax expenditure to be malleable over time? For example, tax depreciation rules have, correctly, increasingly abandoned the principles of economic depreciation in favor of expensing, or the principle that businesses should be able to deduct their investment expenses as incurred. Does this mean accelerated depreciation was once a tax expenditure but is no longer, in which case now the absence of accelerated depreciation is a tax surcharge?

What happens if the result of the JCT general-rule approach conflicts with sound tax theory? As noted, the general rule in an income tax is that interest expense is deductible. Yet JCT continues to show the home mortgage deduction as a tax expenditure.

In each instance, for each question, the JCT analysis flounders and policymakers who rely on it are led astray.

Tax policy would benefit from a sound, rigorous examination of tax expenditures. Policymakers and the public would then have a clearer understanding of the implicit subsidies and penalties granted or imposed through the tax code. This clearer understanding could well facilitate a more productive debate regarding fundamental tax reform. What the tax expenditure debate should not do, however, is become a ruse, a good-policy sheep in a tax-hiking wolf’s clothing. For now, any time the words “reform tax expenditures” are spoken, Americans should hear “tax hike ruse,” and little will be lost in translation.

J. D. Foster, Ph.D., is Norman B. Ture Senior Fellow in the Economics of Fiscal Policy in the Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation.

About the Author

J.D. Foster, Ph.D. Norman B. Ture Senior Fellow in the Economics of Fiscal Policy
Thomas A. Roe Institute for Economic Policy Studies

Show references in this report

[1]J. D. Foster, “Obama Tax Hikes Defended by Myths and Straw Men,” Heritage Foundation Backgrounder No. 2454, August 26, 2010, at

[2]Curtis S. Dubay, “Value-Added Tax: No Easy Fix for the Deficit,” Heritage Foundation WebMemo No. 2772, January 21, 2010, at, and J. D. Foster, “McCain Throws Down the VAT Gauntlet,” The Foundry, Heritage Foundation blog, April 15, 2010, at

[3]Committee for a Responsible Federal Budget, “Let’s Get Specific: Tax Expenditures 2010,” October 14, 2010, at (October 18, 2010); Testimony of Maya MacGuineas, president of the Committee for a Responsible Federal Budget, before the National Commission on Fiscal Responsibility and Reform, July 28, 2010, at (October 14, 2010); and Organisation for Economic Co-operation and Development, “OECD Economic Surveys: United States, September 2010,” 2010, at (October 14, 2010).

[4]The difference between a comprehensive income tax and a consumed-income tax is that a comprehensive income tax base includes all current labor and capital income, thus subjecting savings to multiple layers of taxation. By contrast, a consumed-income tax is applied to all income once and only once.

[5]Joint Committee on Taxation, “Rethinking Tax Expenditures,” May 1, 2008, at (October 18, 2010).

[6] Ibid., p. 7.

[7]The income tax is really a net income tax. For example, businesses are taxed on their net income as opposed to their gross receipts or gross income. In the case of housing, one can rent a house from someone else, in which case the owner pays income tax on the net rental income—monthly rental income net of depreciation, interest, and so forth. Or one can own a home. Either way, whether renter or owner, the house itself generates a stream of services and a stream of net income. The net income is explicit if the home is rented because there are explicit transactions. If the renter is also the homeowner, there are no such transactions and thus the amount of income is unknown and implicit. If one could estimate the income the homeowner earns, then that income would be net imputed income on which income tax would normally be levied.

[8]Joint Committee on Taxation, “Rethinking Tax Expenditures.”